This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrowReprints and Permissions
Right arrow Copyright Information
Right arrow Books from ASM Press
Right arrow MicrobeWorld
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Buchner, J. M.
Right arrow Articles by Karls, A. C.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Buchner, J. M.
Right arrow Articles by Karls, A. C.

 Previous Article  |  Next Article 

Journal of Bacteriology, May 2005, p. 3431-3437, Vol. 187, No. 10
0021-9193/05/$08.00+0     doi:10.1128/JB.187.10.3431-3437.2005
Copyright © 2005, American Society for Microbiology. All Rights Reserved.

Piv Site-Specific Invertase Requires a DEDD Motif Analogous to the Catalytic Center of the RuvC Holliday Junction Resolvases

John M. Buchner, Anne E. Robertson, David J. Poynter,{dagger} Shelby S. Denniston,{ddagger} and Anna C. Karls*

Department of Microbiology, University of Georgia, Athens, Georgia 30602-2605

Received 27 December 2004/ Accepted 7 February 2005


arrow
ABSTRACT
 
Piv, a unique prokaryotic site-specific DNA invertase, is related to transposases of the insertion elements from the IS110/IS492 family and shows no similarity to the site-specific recombinases of the tyrosine- or serine-recombinase families. Piv tertiary structure is predicted to include the RNase H-like fold that typically encompasses the catalytic site of the recombinases or nucleases of the retroviral integrase superfamily, including transposases and RuvC-like Holliday junction resolvases. Analogous to the DDE and DEDD catalytic motifs of transposases and RuvC, respectively, four Piv acidic residues D9, E59, D101, and D104 appear to be positioned appropriately within the RNase H fold to coordinate two divalent metal cations. This suggests mechanistic similarity between site-specific inversion mediated by Piv and transposition or endonucleolytic reactions catalyzed by enzymes of the retroviral integrase superfamily. The role of the DEDD motif in Piv catalytic activity was addressed using Piv variants that are substituted individually or multiply at these acidic residues and assaying for in vivo inversion, intermolecular recombination, and DNA binding activities. The results indicate that all four residues of the DEDD motif are required for Piv catalytic activity. The DEDD residues are not essential for inv recombination site recognition and binding, but this acidic tetrad does appear to contribute to the stability of Piv-inv interactions. On the basis of these results, a working model for Piv-mediated inversion that includes resolution of a Holliday junction is presented.


arrow
INTRODUCTION
 
Piv catalyzes site-specific inversion of a 2.1-kb chromosomal segment encoding type 4 pilin genes of the human and cow eye pathogens Moraxella lacunata and Moraxella bovis, respectively (14, 18, 27). This DNA rearrangement results in phase variation of the type 4 pili (19) (Fig. 1). The invertible segment is bounded by identical 32-bp inverted repeats, invL and invR, where Piv mediates recombination. In addition, Piv interacts with an accessory site, sub1, which may facilitate assembly of an active synaptic complex (32).



View larger version (17K):
[in this window]
[in a new window]
  		FIG. 1. Inversion region on the M. lacunata chromosome. Recombination sites invL and invR are within the coding sequence of the type 4 pilin genes tfpQ and tfpI such that inversion switches the 3' coding sequence of the gene expressed from Ptfp. M. bovis alternately expresses the serologically different pilins, but M. lacunata exhibits a on-off phase variation of TfpQ pili due to a frameshifting 19-bp duplication (black box) in tfpI (18). The invertase Piv, encoded immediately adjacent to the invertible segment, is expressed from Ppiv (10). sub1 is a nonessential Piv binding site (32).

Piv is unique among prokaryotic site-specific DNA invertases because it does not belong to either the tyrosine- or serine-recombinase families. Instead, Piv shares significant amino acid identity and similarity with the transposases of insertion sequences (IS) from the IS110/IS492 family, including three highly conserved amino acid motifs: GDK, KTDDAA, and PSG (14, 24). Piv variants containing alanine or glycine substitutions at multiple positions within each motif are defective for mediating inversion but still bind to the inv and sub sites. The essential role of these conserved motifs in Piv catalysis of DNA inversion suggests that Piv and the transposases of the IS110/IS492 family (Piv/MooV family of DNA recombinases) use similar catalytic mechanisms (25, 31).

Most identified transposases belong to the superfamily of retroviral integrases (IN) and are typified by a conserved DDE catalytic motif. The acidic residues of the DDE motif are separated by variable numbers of amino acids in the transposase primary amino acid sequence but are brought together by protein folding to form a catalytic triad within a RNase H-like structural motif (RNase H-fold) (21). The primary amino acid sequences of Piv/MooV recombinases do not contain a conserved DDE-motif; however, molecular modeling of the tertiary structure for the amino terminal region of Piv predicts an RNase H-fold (31). The completely conserved acidic residues D9 (GDK), E59 (conserved as E or D), D101, and/or D104 (KTDDA) are positioned appropriately in the RNase H-fold to coordinate two divalent metal ions. This potential DED or DEDD catalytic motif for Piv profoundly affects the predicted chemistry for the Piv-mediated site-specific recombination reaction.

Unlike the tyrosine- and serine-recombinases that mediate conservative site-specific recombination by a two-step transesterification reaction, in which the energy of the phosphodiester bond is conserved in a covalent protein-DNA intermediate, the DDE-motif transposases catalyze hydrolysis of the phosphodiester backbone of the DNA substrate and mediate strand transfer by a one-step transesterification reaction. Host enzymes, such as DNA repair or replication functions, are required to complete the transposition process following strand transfer (for reviews, see references 21, 21a, and 32a). The RuvC-related Holliday junction resolvases, which are also members of the retroviral integrase superfamily, utilize a DEDD catalytic tetrad (D7, E66, D138, and D141) within an RNaseH-like fold for coordination of two divalent metal cations (1). Overlapping the crystal structures of the catalytic domains from RuvC and retroviral integrases (IN) shows that three residues of the DEDD motif superimpose on the DDE residues of IN within the RNaseH-fold (21, 34). Like the DDE-motif transposases, RuvC coordinates divalent metal cations to direct hydrolysis of phosphodiester bonds in substrate DNA (reviewed in reference 21). However, RuvC does not catalyze strand transfer following hydrolysis of the substrate DNA, i.e., RuvC cleaves opposing DNA strands in a Holliday junction to resolve the structure and DNA ligase repairs the nicks in the substrate DNA (3, 28). RuvC exhibits sequence specificity, preferring to resolve junctions with the sequence 5'-(A/T)TT(G/C)-3' (6). The sequence 5'-ATTG-3' is near the center of the inv sites, suggesting that if Piv has RuvC-like activity, a Holliday junction generated in recombination between invL and invR sites may be efficiently resolved to give the apparently conservative recombination product.

These intriguing mechanistic predictions for Piv led us to determine the role of the DEDD residues in Piv-mediated inversion and intermolecular recombination. Piv variants with individual substitutions of a small, uncharged amino acid at each potential catalytic residue retain no inversion or intermolecular recombination activity but still bind the inv recombination site. These results indicate that the DEDD motif is part of the Piv active site. Substitutions of residues in the DEDD motif with glutamic acid or aspartic acid, which differ by one carbon in side chain length, result in severely reduced or undetectable Piv catalytic activity, suggesting that there is little flexibility in the spacing of the carboxyl groups within the catalytic tetrad. Interestingly, the variants that retain carboxylates at the predicted catalytic site appear to form more-stable Piv-inv complexes. A working model for Piv-catalyzed recombination is proposed.


arrow
MATERIALS AND METHODS
 
Site-directed mutagenesis and expression of Piv. Piv variants with the amino acid substitutions listed in Table 1 were generated by site-directed mutagenesis (Stratagene QuikChange mutagenesis kit) of piv in pAG630 (31) or in pJMB30, which has the SpeI-to-HindIII fragment containing piv from pAG800.2 (this study) ligated into the same restriction sites of Litmus 29 (New England Biolabs). The mutagenesis oligonucleotides incorporated base pair changes that resulted in the designated missense mutation and, for some, silent mutations that created or eliminated restriction sites in piv for screening mutants. Mutagenized piv genes were subcloned into two different expression vectors for in vivo recombination (pAG800.2) and invL-binding (pJMB52) assays. pAG800.2 contains the BamHI-HindIII fragment from pAG702 (31), encoding piv and its Shine-Dalgarno site, introduced downstream of Ptac in pKH197 (25), and one of two SpeI sites in the pKH197 vector was destroyed by fill-in with Klenow fragment (New England Biolabs). pJMB52 is pKH197 with the BamHI-HindIII region replaced by the BamHI-BsrGI fragment of pACYC-piv1 (14), containing the M. bovis piv promoter (Ppiv) and the first 86 bp of pivmb, and the BsrGI-HindIII fragment of pAG800, encoding the remainder of pivml and invR. Thus, pJMB52 differs from the pAG800.2 expression vector in two aspects: (i) piv is expressed from M. bovis Ppiv, and (ii) piv encodes T21A, one of the five amino acid differences between M. bovis Piv and M. lacunata Piv. The T21A substitution does not affect Piv recombination or DNA binding activity (data not shown).


View this table:
[in this window]
[in a new window]
  		TABLE 1. piv mutations resulting in amino acid substitutions

Two mutant piv expression vectors, pJMB54 and pNull, were created for use as negative controls in the recombination and DNA binding assays. The mutant piv gene inserted into pAG800.2 to give pJMB54 has a 10-bp insertion at the HincII site in piv due to the deletion of the interposon {Omega}Smr/Spr from piv in pMxL5 (18) by BamHI digestion and religation; the insertion causes a frame shift 245 base pairs into piv and results in termination of piv translation after 103 amino acids (Piv{Delta}1). To generate pNull, pJMB52 was digested with BsrGI and HindIII, treated with End-It DNA End-Repair kit (Epicentre), and ligated to restore the vector without the sequence between BsrGI-HindIII. This resulted in a carboxyl-terminal truncation of Piv, leaving only the amino-terminal 28 amino acids (Piv{Delta}2).

All of the mutant piv alleles and the subclones were confirmed by sequencing of the complete piv gene (University of Michigan Biomedical Research Core Facilities [UM-BRCF]).

Expression of wild-type and variant Piv proteins was as follows: cultures of DH5{alpha} with pAG800.2 or pJMB52 carrying the wild-type or mutated piv genes were grown to mid-log phase at 37°C in Luria-Bertani (LB) broth with 50 µg/ml spectinomycin (Sp50), induced with 100 µM IPTG (isopropyl-ß-D-thiogalactopyranoside), and incubated-aerated an additional 2 h, and 1.5 ml was harvested by centrifugation. Cells were resuspended in 270 µl 1x loading buffer (2) and boiled for 5 min, and 90 µl of each sample was electrophoresed on a 12% sodium dodecyl sulfate-polyacrylamide gel. Western blot analysis was performed as described in reference 2 using primary anti-Piv antibody, generated in rabbits against a peptide (CKSDNGIKLTALLKQREHHKRQLIKERTRQE) conjugated to KLH (BIO SYNTHESIS Inc., Lewisville, TX), and secondary anti-rabbit antibody alkaline phosphatase conjugate (Sigma-Aldrich) at 1:1,000 dilutions. Membranes were developed with 5-bromo-4-chloro-3-indolyphosphate-nitroblue tetrazolium SigmaFast tablets (Sigma-Aldrich).

In vivo inversion assays. The inversion substrates, pMxL90 (Q-orientation substrate) and pMxL100 (I-orientation substrate), were derived from pMxL5 and pMxL6 (18), respectively, by replacing the {Omega}Smr/Spr cassette that interrupts piv with the {Omega}Cmr (chloramphenicol resistance) interposon from pHP45{Omega}-Cm (8). Chemically competent Escherichia coli DH5{alpha} cells containing either pMxL90 or pMxL100 were transformed with pAG800.2 or a Piv variant expression vector. Transformants were plated on LB agar containing Sp50, Cm34, and 50 µM IPTG. After 24 to 36 h incubation at 37°C individual colonies were inoculated into LB broth-Sp50-Cm34 and incubated-aerated at 37°C for approximately 24 h. The inversion substrate and expression vector were isolated using Wizard Plus Miniprep kits (Promega). Two different methods were used to detect inversion of the invL/invR-flanked segment on the substrate plasmid: (i) 200 to 300 ng of the isolated plasmid DNA was digested with BsrGI, electrophoresed on 0.6% agarose gels, stained with ethidium bromide (EtBr), and imaged with a Bio-Rad Fluor-S Multi-imager (Bio-Rad Quantity One software was used to determine relative intensities of bands to estimate the percentage of inversion); (ii) 20 to 50 ng of the isolated DNA was used in a three-primer PCR mixture containing 1x KlenTaq DV ReadyMix (Sigma Aldrich) and 200 nM of each primer (T7 [5'-GTAATACGACTCACTATAGGGC-3'], IAR [5'-CTAACCATCAGCTATGCCGTTATTC-3'], and IAF5 [5'-CATGATATGCTGCTTGACCCCAACC-3']), and 5 µl of each PCR mixture (25 µl total volume, which was cycled in a Bio-Rad i-cycler for 1 cycle at 95°C for 60 s, 25 cycles at 95°C for 45 s, 61°C for 30 s, and 72°C for 45 s, and 1 cycle at 72°C for 7 min) was electrophoresed on a 1.2% agarose gel and visualized as described above. Each Piv variant was tested for inversion from the Q-to-I and I-to-Q orientations from at least three independent transformants of DH5{alpha} containing pMxL90 or pMxL100.

Intermolecular recombination assays. Intermolecular recombination between invR on pAG800.2 and invL on pMxL90 or pMxL100 was assayed by PCR: 25-µl reactions contained 1x Taq buffer B (Fisher), 1.5 mM MgCl2, 1 µM of each primer (CAP [5'-GGCTGGCTTTTTCTTGTTATCGC-3'] and BLU [5'-GGGTTATTGTCTCATGAGCGG-3']), 0.2 µM deoxynucleotide triphosphates, 10 to 25 ng of DNA isolated from the inversion assay transformants (described above), and 1.5 units Taq DNA polymerase (Fisher). Cycling conditions for the reactions were as follows: 1 cycle at 95°C for 60 s, 25 cycles at 95°C for 30 s, 64°C for 30 s, and 68°C for 36 s, and 1 cycle at 68°C for 36 s; products were electrophoresed and imaged as described above. Selected products were inserted into pCR2.1 (Invitrogen) and sequenced (UM-BRCF) to determine the site of DNA exchange. These assays were repeated with Pfu (Stratagene) instead of Taq polymerase with the following changes: the reaction mix contained 1x Pfu buffer (Stratagene), 0.5 µM of each primer, 0.25 µM deoxynucleotide triphosphates, and 1.9 units Pfu DNA polymerase, and the cycling conditions were changed to 1 cycle at 95°C for 60 s, 25 cycles at 95°C for 30 s, 64°C for 30 s, and 72°C for 2 min 22 s, and 1 cycle at 72°C for 10 min. Products were analyzed as described above. The DNA from at least three independent inversion assays for each Piv variant was analyzed for intermolecular recombination products.

In vivo DNA binding assays. To measure in vivo binding of Piv to inv, the plasmid system of Elledge and Davis (7) was utilized. A double-stranded oligonucleotide encoding the inv sequence (top strand, 5'-GCCATTATTGGTATCCTAGCTGCAATCGCT-3') was inserted into the SmaI site at +1 of the conII promoter sequence in pNN396 (7). The NotI-HindIII restriction fragment, containing the conII promoter and the inv sequence, was then ligated into the same restriction sites on pNN387 upstream of lacZ lacY to create pAR110. DH5{alpha} containing pAR110 was transformed with Piv wild-type or Piv variant expression vectors; cultures of the transformants in LB-Sp50-Cm34 were grown to an optical density of 0.7 to 0.9. Cells were harvested from two separate 1-ml aliquots by centrifugation: one pellet was resuspended in 1 ml Z buffer to use in ß-galactosidase activity assays (20), and the other was resuspended in 1 ml Z buffer without ß-mercaptoethanol to measure total protein concentration with a bicinchoninic protein assay (Pierce). In both assays the cells were lysed by addition of two drops of chloroform and 1 drop of 0.1% sodium dodecyl sulfate. All samples were assayed in duplicate for 1:2 and 1:10 dilutions in Z buffer. Each Piv variant was assayed from at least three independent transformants, and the wild-type Piv and Piv{Delta}2 were assayed from eight independent transformants. The ß-galactosidase activity is expressed as Miller units (20) per µg protein in each sample.


arrow
RESULTS
 
DEDD motif residues are all required for Piv inversion activity. Previous mutational analysis of the conserved amino acid motifs from the Piv/MooV recombinase family utilized multiple substitutions within each conserved motif in Piv, leading to conclusions about the contributions of the motif as a whole in Piv catalysis of inversion (31). To address the role of individual amino acids within the conserved motifs, focusing on those residues that are predicted to constitute a DED or DEDD catalytic motif, Piv variants were generated that are singularly substituted at D9, E59, D101, and D104 with alanine, glycine, aspartic acid, or glutamic acid (Fig. 2). All of these Piv variants were expressed at levels comparable to wild-type Piv, as measured by Western blot analysis (Fig. 3), suggesting that the variant proteins can fold appropriately. In contrast, none of these Piv variants gave detectable levels of in vivo inversion activity in an assay that directly measured inversion of the pilin segment on a plasmid (Fig. 4). The limit of detection in this assay is ~1% inversion; wild-type Piv exhibits ~30% inversion of the pilin segment on the plasmid substrate. The results for switching from the tpfQ (Q) to tfpI (I) orientation for the invertible segment is shown in Fig. 4; assaying inversion from I to Q yielded identical results (data not shown). These results support the proposal that the DEDD residues comprise a catalytic tetrad within Piv.



View larger version (17K):
[in this window]
[in a new window]
  		FIG. 2. Substitutions in Piv. The predicted catalytic residues within the amino terminal 160 amino acids of Piv are highlighted by gray boxes, and the nonconserved glutamic acid residues that were also targeted for mutagenesis are in open boxes. The substituted residues are indicated in gray type below the wild-type amino acid.



View larger version (38K):
[in this window]
[in a new window]
  		FIG. 3. Expression of Piv variants in the strain used for inversion assays. Mid-log cultures of DH5{alpha}, or DH5{alpha} carrying pAG800.2 encoding the wild-type (wt) or mutated piv genes (the substitutions are indicated above each lane) were induced with 100 µM IPTG, and at 2 h postinduction, cells were harvested and lysed and the proteins were fractionated by electrophoresis on 12% sodium dodecyl sulfate-polyacrylamide gels. The Western blot of this gel, utilizing anti-Piv antisera as primary antibody, is shown. Piv is marked by an arrow.



View larger version (62K):
[in this window]
[in a new window]
  		FIG. 4. In vivo inversion activity of Piv variants substituted within the DEDD motif and at nonconserved glutamate residues. DH5{alpha}, containing the inversion substrate pMxL90, was transformed with pAG800.2-derived expression vectors, encoding wild-type Piv or the variants with the indicated substitutions, and Piv expression was induced with 50 µM IPTG. Plasmid DNA was isolated from overnight cultures of individual transformants, and inversion of the type 4 pilin segment on pMxL90 was determined by digestion with BsrGI. Digest products were electrophoresed on a 0.6% agarose gel and stained with EtBr (inverted image is shown). The starting Q orientation of the invertible segment on pMxL90 (Q) yields unique 6.5- and 2.6-kb fragments (pMxL90 alone); inversion to the I orientation gives unique 5.2- and 3.9-kb fragments (pMxL100 alone); a 2.5-kb fragment is common to both. The expression vectors contain only one BsrGI site, giving a single 7-kb fragment (pAG800.2 alone).

Each of the three glutamic acid residues that are 23, 34, and 43 residues beyond D104 is spaced appropriately in the primary amino acid sequence to complete a DDE motif in which the first two residues consist of D9 and D101/D104. However, this potential catalytic motif is not predicted in the molecular modeling of Piv and is not conserved among the Piv/MooV family members. Substitution of E127, E138, and E147 individually with alanine or glycine did not significantly reduce Piv expression (Fig. 3) or inversion activity (Fig. 4), indicating that these acidic residues do not play an essential role in Piv-mediated recombination.

Piv variants with carboxylate substitutions at E59 or D101 retain low levels of catalytic activity. While D9, D101, and D104 are completely conserved in the Piv/MooV recombinases, a subgroup of the family has an aspartic acid at the E59 position. Therefore, the possibility that DDDD could serve equally well as DEDD to form a catalytic tetrad in Piv was tested with the Piv variant E59D. Although this variant did not mediate detectable inversion in the assay described above (Fig. 4), a PCR-based assay for intermolecular recombination revealed that Piv E59D does catalyze recombination between invL on the inversion substrate and invR on the Piv expression vector, albeit at a low level (Fig. 5). Therefore, to investigate whether Piv E59D mediates inversion but at a level below the detection limits of the plasmid restriction digest assay, PCR was utilized to detect inversion products in the plasmid DNA isolated from the in vivo inversion experiment. As seen in Fig. 6, there is inversion product from the Piv E59D assay; this product was detected at low levels in five out of six independent in vivo inversion assays (3 of 3 Q-to-I assays and 2 of 3 I-to-Q assays; data not shown). However, these PCR assays also revealed that Piv-independent inversion occurs at very low levels (Fig. 6). PCR-mediated recombination or template switching could explain the Piv-independent inversion products that are detected by PCR, but it has been demonstrated that Pfu DNA polymerase does not mediate recombination under normal PCR conditions (29), and the same results are obtained with Taq and Pfu DNA polymerases (data not shown). Replication-dependent, recombination-independent inversion-dimerization of plasmids containing long inverted repeats has been characterized previously (5, 17) and likely explains the observed levels of Piv- and RecA-independent inversion.



View larger version (45K):
[in this window]
[in a new window]
  		FIG. 5. In vivo intermolecular recombination mediated by Piv variants substituted at acidic residues. DNA from the inversion assays with Piv variants described in Fig. 4 was assayed for intermolecular recombination between invL on pMxL90 and invR on the pAG800.2-derived vectors. The new DNA junction was detected by PCR using primers, designated P1 and P2, that anneal to sequence unique to pAG800.2 and pMxL90, respectively. PCR products were electrophoresed on a 1.2% agarose gel and stained with EtBr (inverted image is shown). The 1,073-bp PCR product was sequenced from selected reactions to confirm that recombination occurred within the inv sequences.



View larger version (59K):
[in this window]
[in a new window]
  		FIG. 6. Detection of low level in vivo inversion activity of Piv variants. DNA template utilized in the recombination assays described in Fig. 4 and 5 was used in a three-primer PCR inversion assay. A primer that anneals to tfpB sequence on pMxL90 pairs with one of two primers that anneal to sequence flanking the invertible segment to yield a 981- or a 811-bp PCR product when the invertible segment is in the "Q" or "I" orientation, respectively. The PCR products were electrophoresed on a 1.2% agarose gel and stained with EtBr (inverted image is shown).

Judging on the basis of the molecular modeling of Piv, either D101 or D104 could participate in a catalytic DED triad, similarly to the DDE motif of transposases. The substitution of either carboxylate with a small, uncharged amino acid results in loss of Piv catalytic activity as measured by both PCR-based recombination assays (Fig. 5 and 6). This result suggests that both D101 and D104 are required as part of a catalytic DEDD tetrad, as seen for the RuvC-related resolvases (30). Substitution of the D101 position with another carboxylate (D101E) is partially tolerated, yielding low levels of recombination activity, while replacement of D104 with glutamic acid results in loss of all recombination activity (Fig. 5 and 6). In addition, rotating the position of these carboxylates with E59 to create Piv variants with DDED (E59D/D101E) and DDDE (E59D/D104E) motifs resulted in stable (Fig. 3) but catalytically inactive (Fig. 4, 5, and 6) proteins, indicating that the spatial positions of the carboxyl group of D101 and D104 are important for Piv catalytic activity.

DEDD residues are not essential for binding the recombination sites. Although in vitro binding assays have demonstrated that Piv binds weakly to the inv recombination sites (32), Piv interactions with invL can be assessed in vivo using a transcription repression system (7, 31). Expression of lacZ from a constitutive promoter, PconII (7), on a single-copy vector is controlled by Piv binding to invL sequence that functions as an operator site. Thus, a higher level of ß-galactosidase activity reflects a lower level of Piv binding to the invL site. The basal level of ß-galactosidase activity is determined with binding-defective Piv{Delta}2 expressed from pNull (Fig. 7). Wild-type Piv and Piv variants with acidic amino acid substitutions for the predicted catalytic residues (E59D, D101E, and D104E) bound to invL, resulting in a 50 to 57% reduction in ß-galactosidase activity relative to the basal level (Fig. 7). Interestingly, variants with alanine and glycine substitutions of DEDD-motif residues bound invL but exhibited lower repression levels; binding of D9A, D9G, E59A, D101A, D101G, and D104A Piv variants reduced ß-galactosidase activity by 24, 32, 26, 35, 41, and 38%, respectively (Fig. 7). Substitution of the nonconserved residues with alanine or glycine did not affect Piv binding to invL (55 to 62% reduction in ß-galactosidase activity; Fig. 7). The DDED and DDDE double-substitution variants reduced ß-galactosidase activity by 41 and 52%, respectively (Fig. 7). These results indicate that the predicted catalytic residues are not essential for recognizing and binding the recombination sites, but the DEDD motif may contribute to protein conformation or protein-DNA interactions that stabilize interactions of Piv with the inv sites.



View larger version (33K):
[in this window]
[in a new window]
  		FIG. 7. In vivo DNA binding activity of Piv variants. The invL recombination site, introduced at the +1 position relative to a constitutive promoter, acts as an operator sequence controlling expression of lacZ on a single-copy plasmid, pAR110 (7). Piv wild-type (wt) and Piv variant (substitutions are indicated) expression vectors were transformed into the DH5{alpha} strain containing pAR110, and the transformants were assayed for ß-galactosidase activity. The ß-galactosidase activity is indicated as Miller units per µg protein.


arrow
DISCUSSION
 
An acidic residue tetrad, DEDD, is required for Piv catalysis of recombination. We previously hypothesized, on the basis of molecular modeling of Piv tertiary structure and mutagenesis of the conserved amino acid motifs from the Piv/MooV family, that Piv utilizes a DED motif that is equivalent to the DDE motif of transposases (31). While the model clearly predicted D9 and E59 to be the first two catalytic residues, it was possible that either D101 or D104 was the third residue of the catalytic motif. The substitution studies herein show that D101 and D104, in addition to D9 and E59, are essential residues in the Piv active site. D9A, D9G, E59A, D101A, D101G, D104A, and D104E substitutions resulted in Piv variants that are completely defective for in vivo inversion and intermolecular recombination. Piv E59D and Piv D101E exhibit very low levels of recombination, and generating a DDE-, versus DED-, motif in the linear amino acid sequence of Piv (Piv E59D/D101E and Piv E59D/D104E) did not yield an active recombinase. These results are consistent with substitution analyses of the DDE-motif in Tn5 Tnp and the DEDD-motif in the Holliday junction resolvase RuvC (26, 28). These studies showed that substitution of catalytic residues with nonacidic amino acids eliminates recombination activity, but interchanging aspartic acid and glutamic acid residues at catalytic positions (other than the first aspartic acid) is tolerated to give low levels of in vitro catalytic activity. Peterson and Reznikoff (26) concluded that small shifts in the spatial positions of the carboxyl groups within the Tn5 Tnp active site decrease the efficiency with which two divalent metal cations are coordinated for catalytic reactions (16). It is likely that switching of the carboxylate residues in the DEDD tetrad of Piv also alters optimal spacing of the carboxyl groups involved in coordination of ions at the catalytic site.

Role for catalytic residues in Piv interactions with the recombination sites. The results of in vivo binding assays with DEDD-motif Piv variants that were individually substituted at each position of the motif demonstrate that these acidic residues are not essential for Piv binding to the recombination site. However, the nature of substitutions at these positions did influence in vivo binding. As measured by repressor-operator activity of Piv/invL complexes on lacZ expression from PconII, replacing E59, D101, and D104 with acidic residues gave the same level of repression that was obtained with wild-type Piv; however, substitutions with alanine or glycine reduced Piv-mediated repression by 14 to 24%. Substitution of the nonconserved residues E127, E138, and E147 with alanine or glycine did not affect binding of Piv to invL in these in vivo assays. These results indicate that the catalytic residues may play a role in stabilizing the interactions of Piv with the recombination site. A similar role for the DDE motif of Tn10 Tnp in target DNA binding has been suggested by Junop and Haniford (11). Although the DDE motif is not required for target site selection by Tn10 Tnp, the DDE residues are needed for capture of a suboptimal target site (11). The function of these acidic residues in stabilizing recombinase-target interactions involves coordination of the divalent metal ions, which could create bridging contacts between the transposase and the substrate DNA or might stabilize the optimal transposase conformation for DNA binding (11).

A working model for Piv-mediated recombination. Our results demonstrate that all four residues of the DEDD motif in Piv are required for catalysis of inversion and intermolecular recombination. Thus, the catalytic domain of Piv probably more directly resembles that of RuvC than that of the DDE-motif transposases. But what does this imply about the mechanism for Piv-mediated recombination? The catalytic domain of RuvC is remarkably similar to that of the DDE-motif transposases (34). The DEDD and DDE catalytic residues of RuvC and transposases, respectively, are positioned within the RNase H-fold to coordinate divalent metal cations that direct hydrolysis of the phosphodiester bond in substrate DNA, generating a free 3' OH end (reviewed in reference 21). A primary difference between the activities of RuvC and transposases is that RuvC does not catalyze strand transfer following hydrolytic cleavage of the Holliday junction, while transposases utilize the 3' OH of cleaved donor DNA as the nucleophile to attack the phosphodiester bond of target DNA in strand transfer. To accomplish this polynucleotidyl transfer reaction transposases retain the 3' end of the cleaved donor DNA in the active site, blocking the entry of another nucleophile, and bind the target DNA within the same active site to serve as a substrate for the one-step transesterification reaction (12). The third aspartic acid in the RuvC catalytic site may preclude binding of both the cleaved 3' OH end and a target DNA strand within the active site by immobilizing the first nucleophile and/or the 3' OH end, thus preventing the adjustments needed to accommodate the target DNA strand. Alternatively, the absence of a RuvC-mediated strand transfer may simply reflect that there is no target strand available for binding to the catalytic site within the constrained structure of the RuvC-bound Holliday junction.

Thus, in our working model for Piv-mediated inversion we utilize both hydrolysis-strand transfer and endonucleolytic activities for the DEDD active site in the reaction leading to site-specific recombination. The model (Fig. 8) shows Piv bound as a dimer to the synapsed recombination sites (it is also possible that dimers are bound to each inv site and synapse, forming a tetramer). Binding and DNA cleavage-strand transfer induces a conformational change in the Piv-DNA complex so that new recombinase-DNA contacts promote formation of a specific Holliday junction structure, an activity demonstrated for RuvC (4) and other Holliday junction resolvases (9, 33). The active sites of the Piv dimer are now positioned to cleave the unexchanged strands. DNA ligase may repair the original nick before or after the RuvC-like activity that leads to resolution of the junction and is also required for repair of the nicks in the recombined DNA strands. It is arguable that the Piv-nicked-DNA complex is stable prior to DNA ligase activity on the basis of the robust nature of some intermediates in transposition systems, such as the bacteriophage Mu transpososome which needs ClpX unfolding activity to disassemble (remodel) the strand transfer complex (reviewed in reference 22).



View larger version (42K):
[in this window]
[in a new window]
  		FIG. 8. A working model for Piv-mediated site-specific DNA inversion. The synapsed invL (black lines) and invR (gray lines) are shown bound by a Piv dimer (gray circles). Piv mediated-DNA hydrolysis at one recombination site (depicted as invL) (a) and strand transfer (b) lead to formation of a Holliday junction structure. Repositioned Piv catalytic sites now cleave the outer strands of the junction (c), and host DNA ligase activity repairs the nicks (d).

The initial cleavage and strand transfer steps of this working model are similar to those leading to branched DNA intermediates in transposition of IS911 (15) and IS30 (13) adjacent to a copy of the IS inverted repeat. It has been demonstrated that the IS30 Tnp can mediate inversion of DNA between two dyad symmetric sites containing IS30 inverted repeats (13). In the models for recombination mediated by the IS911 and IS30 Tnp (15, 23) host functions process the branch intermediates set up by transposase-mediated strand transfer. In our model for Piv-mediated inversion Piv promotes formation of the Holliday junction and Piv, rather than host functions, specifically resolves the junction close to the original strand transfer and without significant heteroduplex DNA formation. We are currently testing these predictions of the model utilizing both genetic and biochemical approaches.


arrow
ACKNOWLEDGMENTS
 
This work was supported by National Institutes of Health grant GM49794 (to A.C.K.).

We thank Ellen Neidle for critical review of the manuscript, Gordon Churchward and Mick Chandler for helpful discussions, Weiping Fu for constructing Piv E147A expression plasmid, Laurie Barsh Dirksen for generating pAG800 and pAG801, and How-Yi Chang and Valerie Gordon (National Science Foundation Research Experiences for Undergraduates Summer Program at University of Georgia participant) for their efforts on this project.


arrow
FOOTNOTES
 
* Corresponding author. Mailing address: Department of Microbiology, University of Georgia, Biological Sciences Building, Cedar St., Athens, GA 30602-2605. Phone: (706) 583-0822. Fax: (706) 542-2674. E-mail: akarls{at}uga.edu. Back

{dagger} Present address: Wake Forest University School of Medicine, Winston-Salem, NC. Back

{ddagger} Present address: University of Tennessee Health Science Center, College of Pharmacy, Memphis, TN. Back


arrow
REFERENCES
 
    1
  1. Ariyoshi, M., D. G. Vassylyev, H. Iwasaki, H. Nakamura, H. Shinagawa, and K. Morikawa. 1994. Atomic structure of the RuvC resolvase: a Holliday junction-specific endonuclease from E. coli. Cell 78:1063-1072.[CrossRef][Medline]
    2. 2
  2. Ausubel, F. M., R. Brent, R. E. Kingston, D. D. Moore, J. G. Seidman, J. A. Smith, and K. Struhl. 1991. Current protocols in molecular biology. John Wiley & Sons, Inc., New York, N.Y.
    3. 3
  3. Bennett, R. J., H. J. Dunderdale, and S. C. West. 1993. Resolution of Holliday junctions by RuvC resolvase: cleavage specificity and DNA distortion. Cell 74:1021-1031.[CrossRef][Medline]
    4. 4
  4. Bennett, R. J., and S. C. West. 1995. Structural analysis of the RuvC-Holliday junction complex reveals an unfolded junction. J. Mol. Biol. 252:213-226.[CrossRef][Medline]
    5. 5
  5. Bi, X., and L. F. Liu. 1996. DNA rearrangement mediated by inverted repeats. Proc. Natl. Acad. Sci. USA 93:819-823.[Abstract/Free Full Text]
    6. 6
  6. Eggleston, A. K., and S. C. West. 2000. Cleavage of Holliday junctions by the Escherichia coli RuvABC complex. J. Biol. Chem. 275:26467-26476.[Abstract/Free Full Text]
    7. 7
  7. Elledge, S. J., and R. W. Davis. 1989. Position and density effects on repression by stationary and mobile DNA-binding proteins. Genes Dev. 3:185-197.[Abstract/Free Full Text]
    8. 8
  8. Fellay, R., J. Frey, and H. Krisch. 1987. Interposon mutagenesis of soil and water bacteria: a family of DNA fragments designed for in vitro insertional mutagenesis of gram-negative bacteria. Gene 52:147-154.[CrossRef][Medline]
    9. 9
  9. Giraud-Panis, M. J., and D. M. Lilley. 1998. Structural recognition and distortion by the DNA junction-resolving enzyme RusA. J. Mol. Biol. 278:117-133.[CrossRef][Medline]
    10. 10
  10. Heinrich, D. W., and A. C. Glasgow. 1997. Transcriptional regulation of type 4 pilin genes and the site-specific recombinase gene, piv, in Moraxella lacunata and Moraxella bovis. J. Bacteriol. 179:7298-7305.[Abstract/Free Full Text]
    11. 11
  11. Junop, M. S., and D. B. Haniford. 1997. Factors responsible for target site selection in Tn10 transposition: a role for the DDE motif in target DNA capture. EMBO J. 16:2646-2655.[CrossRef][Medline]
    12. 12
  12. Kennedy, A. K., D. B. Haniford, and K. Mizuuchi. 2000. Single active site catalysis of the successive phosphoryl transfer steps by DNA transposases: insights from phosphorothioate stereoselectivity. Cell 101:295-305.[CrossRef][Medline]
    13. 13
  13. Kiss, J., M. Szabo, and F. Olasz. 2003. Site-specific recombination by the DDE family member mobile element IS30 transposase. Proc. Natl. Acad. Sci. USA 100:15000-15005.[Abstract/Free Full Text]
    14. 14
  14. Lenich, A. G., and A. C. Glasgow. 1994. Amino acid sequence homology between Piv, an essential protein in site-specific DNA inversion in Moraxella lacunata, and transposases of an unusual family of insertion elements. J. Bacteriol. 176:4160-4164.[Abstract/Free Full Text]
    15. 15
  15. Loot, C., C. Turlan, and M. Chandler. 2004. Host processing of branched DNA intermediates is involved in targeted transposition of IS911. Mol. Microbiol. 51:385-393.[CrossRef][Medline]
    16. 16
  16. Lovell, S., I. Y. Goryshin, W. R. Reznikoff, and I. Rayment. 2002. Two-metal active site binding of a Tn5 transposase synaptic complex. Nat. Struct. Biol. 9:278-281.[CrossRef][Medline]
    17. 17
  17. Lyu, Y. L., C. T. Lin, and L. F. Liu. 1999. Inversion/dimerization of plasmids mediated by inverted repeats. J. Mol. Biol. 285:1485-1501.[CrossRef][Medline]
    18. 18
  18. Marrs, C. F., F. W. Rozsa, M. Hackel, S. P. Stevens, and A. C. Glasgow. 1990. Identification, cloning, and sequencing of piv, a new gene involved in inverting the pilin genes of Moraxella lacunata. J. Bacteriol. 172:4370-4377.[Abstract/Free Full Text]
    19. 19
  19. Marrs, C. F., W. W. Ruehl, G. K. Schoolnik, and S. Falkow. 1988. Pilin-gene phase variation of Moraxella bovis is caused by an inversion of the pilin genes. J. Bacteriol. 170:3032-3039.[Abstract/Free Full Text]
    20. 20
  20. Miller, J. H. 1992. A short course in bacterial genetics. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, N.Y.
    21. 21
  21. Mizuuchi, K. 1997. Polynucleotidyl transfer reactions in site-specific DNA recombination. Genes Cells 2:1-12.[Abstract]
    22. 21
  22. Nagy, Z., and M. Chandler. 2004. Regulation of transposition in bacteria. Res. Microbiol. 155:387-398.[Medline]
    23. 22
  23. Nakai, H., V. Doseeva, and J. M. Jones. 2001. Handoff from recombinase to replisome: insights from transposition. Proc. Natl. Acad. Sci. USA 98:8247-8254.[Abstract/Free Full Text]
    24. 23
  24. Olasz, F., T. Fischer, M. Szabo, Z. Nagy, and J. Kiss. 2003. Gene conversion in transposition of Escherichia coli element IS30. J. Mol. Biol. 334:967-978.[CrossRef][Medline]
    25. 24
  25. Partridge, S. R., and R. M. Hall. 2003. The IS1111 family members IS4321 and IS5075 have subterminal inverted repeats and target the terminal inverted repeats of Tn21 family transposons. J. Bacteriol. 185:6371-6384.[Abstract/Free Full Text]
    26. 25
  26. Perkins-Balding, D., G. Duval-Valentin, and A. C. Glasgow. 1999. Excision of IS492 requires flanking target sequences and results in circle formation in Pseudoalteromonas atlantica. J. Bacteriol. 181:4937-4948.[Abstract/Free Full Text]
    27. 26
  27. Peterson, G., and W. Reznikoff. 2003. Tn5 transposase active site mutations suggest position of donor backbone DNA in synaptic complex. J. Biol. Chem. 278:1904-1909.[Abstract/Free Full Text]
    28. 27
  28. Rozsa, F. W., and C. F. Marrs. 1991. Interesting sequence differences between the pilin gene inversion regions of Moraxella lacunata ATCC 17956 and Moraxella bovis Epp63. J. Bacteriol. 173:4000-4006.[Abstract/Free Full Text]
    29. 28
  29. Saito, A., H. Iwasaki, M. Ariyoshi, K. Morikawa, and H. Shinagawa. 1995. Identification of four acidic amino acids that constitute the catalytic center of the RuvC Holliday junction resolvase. Proc. Natl. Acad. Sci. USA 92:7470-7474.[Abstract/Free Full Text]
    30. 29
  30. Shafikhani, S. 2002. Factors affecting PCR-mediated recombination. Environ. Microbiol. 4:482-486.[CrossRef][Medline]
    31. 30
  31. Sharples, G. J. 2001. The X philes: structure-specific endonucleases that resolve Holliday junctions. Mol. Microbiol. 39:823-834.[CrossRef][Medline]
    32. 31
  32. Tobiason, D. M., J. M. Buchner, W. H. Thiel, K. M. Gernert, and A. C. Karls. 2001. Conserved amino acid motifs from the novel Piv/MooV family of transposases and site-specific recombinases are required for catalysis of DNA inversion by Piv. Mol. Microbiol. 39:641-651.[CrossRef][Medline]
    33. 32
  33. Tobiason, D. M., A. G. Lenich, and A. C. Glasgow. 1999. Multiple DNA binding activities of the novel site-specific recombinase, Piv, from Moraxella lacunata. J. Biol. Chem. 274:9698-9706.[Abstract/Free Full Text]
    34. 32
  34. Turlan, C., and M. Chandler. 2000. Playing second fiddle: second-strand processing and liberation of transposable elements from donor DNA. Trends Microbiol. 8:268-274.[CrossRef][Medline]
    35. 33
  35. White, M. F., and D. M. Lilley. 1997. The resolving enzyme CCE1 of yeast opens the structure of the four-way DNA junction. J. Mol. Biol. 266:122-134.[CrossRef][Medline]
    36. 34
  36. Yang, W., and T. A. Steitz. 1995. Recombining the structures of HIV integrase, RuvC and RNase H. Structure 3:131-134.[Medline]

Journal of Bacteriology, May 2005, p. 3431-3437, Vol. 187, No. 10
0021-9193/05/$08.00+0     doi:10.1128/JB.187.10.3431-3437.2005
Copyright © 2005, American Society for Microbiology. All Rights Reserved.



This article has been cited by other articles:

  • Tetu, S. G., Holmes, A. J. (2008). A Family of Insertion Sequences That Impacts Integrons by Specific Targeting of Gene Cassette Recombination Sites, the IS1111-attC Group. J. Bacteriol. 190: 4959-4970 [Abstract] [Full Text]  
  • Higgins, B. P., Carpenter, C. D., Karls, A. C. (2007). Chromosomal context directs high-frequency precise excision of IS492 in Pseudoalteromonas atlantica. Proc. Natl. Acad. Sci. USA 104: 1901-1906 [Abstract] [Full Text]  
  • Prosseda, G., Latella, M. C., Casalino, M., Nicoletti, M., Michienzi, S., Colonna, B. (2006). Plasticity of the Pjunc Promoter of ISEc11, a New Insertion Sequence of the IS1111 Family. J. Bacteriol. 188: 4681-4689 [Abstract] [Full Text]  
  • Ramos-Gonzalez, M. I., Campos, M. J., Ramos, J. L., Espinosa-Urgel, M. (2006). Characterization of the Pseudomonas putida Mobile Genetic Element ISPpu10: an Occupant of Repetitive Extragenic Palindromic Sequences. J. Bacteriol. 188: 37-44 [Abstract] [Full Text]  

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrowReprints and Permissions
Right arrow Copyright Information
Right arrow Books from ASM Press
Right arrow MicrobeWorld
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Buchner, J. M.
Right arrow Articles by Karls, A. C.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Buchner, J. M.
Right arrow Articles by Karls, A. C.

Copyright © 2009 by the American Society for Microbiology.   For an alternate route to Journals.ASM.org, visit: http://intl-journals.asm.org | More Info»